How to Create the "Pull" for Patent Licensing?

OK, so here we go again. Sticking to my comfort zone, IP/tech licensing, still this time around. But do warn in advance that the next time around we will shift gears and dwell deep into the societal and legal impacts of commonly available DNA sequencing services, such as 23andme.com. Also trying to keep the frequency and cadence of these to about once every month. Really out of respect for others. Have no intention to write about things, where I do not believe I can add valuable insights and so on.

Patent Licenses as Products

Patent licenses are not your normal products. In fact they are extremely unusual as products. I have heard patent licenses be characterized as offering "peace of mind", "insurance against being sued", "freedom of operation", "access to technology" and many other things, including "relief from fear and pain". People, me included, usually struggle for at least the first 5-10 years in licensing before really getting the context and product understanding right. Just takes time. Hope this article helps to expedite that.

Now, let me illustrate through a few examples. First, let us think of someone who wants to practice a patent. The whole idea with a patent is that there is this theoretic trade where the inventor discloses his idea, and how to practice it, in exchange for being granted an exclusive right to invention for the 20 year patent term. Sounds poetic and beautiful, does it not? Well, if the system works at all, then it also means that anyone well versed in the relevant technology, or as the patent geeks would say "a person skilled in the art", should be able to work and practice the invention based on the teaching of the patent. So, really, from a purely knowledge and skills perspective, you generally should not need patent licenses to be able to do something. Quite the opposite, you just need the patent document for your resident techies in R&D.

Second, let's think of the following scenario: I want to make smartphones. I really, really, really do. For money, of course. All the smartphone patent holders out there can only grant licenses (i.e., they can agree to refrain from litigating against me) under their own patents only. But generally not under anyone else's patents (absent some rare and limited setups such as the right to sublicense someone else's patents). Anyhow, there is nobody who can license me all the patents I need to make a smartphone. Just is not. And never will be. This is sometimes expressed as patents being negative rights to exclude the practicing of the patented invention. And so they are. The patent holder can only agree to refrain from saying NO under its own patents. Nothing more. Importantly, the patent holder can neither say NO nor YES under patents held by others. Put differently, get license from one guy and there are still dozens in the line. It never ends, nor should it.

Finally, let's be positive such that, despite all of this, someone, maybe in the country they themselves call the Middle Kingdom or the Center of the Universe, (depending on how literal you get with the translations for calling that country) has done the following. They have actually taken a patent license from someone. Drum roll !!! Now, let's say that times get rough for them. And they must stop paying for one of the following: salaries for the factory personnel, lease for the factory, components, electricity, taxes or the patent license. The last two they can stop without immediate problems for production. With taxes, there may be grave consequences like imprisonment or even something graver than that. But if you fail to pay under the patent license and you then you typically risk getting a letter requesting royalty reports and audit, and even that usually with some delay. Then you stall and fight the royalty audit in courts or arbitration for a few years. Then you stall the audit. And then you fight another litigation to stall having to pay the omitted royalties. Easily 5-6 years or even more of wiggle room. Which the taxman will not allow you. And not paying for electricity, components, factory lease and salaries means instant death for your business. And not paying taxes gets you at least imprisoned. Not cool. So what you do? Think about it. Breach of contract, but the consequences hit with a delay, unless the licensor has been savvy and built in separate protections against that. More on that in some other post.

OK, So the Product is Strange, So How Do We Sell It?

By now, I hope to have convinced you that the product is strange. That does not mean that patent licensing is a bad business. Quite the opposite. Done right, and with a little luck, it is really one of the best, if not the best business, in the world. Scales like crazy when all the stars align. Well, it is hard to compete with some other businesses like, say, Justin Bieber and various outlawed businesses. Anyhow, with patent licenses, and I have been involved in hundreds of them and been part of several amazing successes, you really must pay attention to creating the "pull" needed for there to be demand for your rather esoteric product. So how do you do it?

Marshall's Three Approaches

Everyone in patent licensing likes its grand daddy Marshall Phelps, the former head honcho first at IBM and then later at Microsoft IP licensing. Well, he has been involved in some rather odd projects more recently, such as the failed IPXI patent license "stock exchange", but he does have a great reputation. I have only met him a few times. Anyhow, Marshall has long ago categorized the ways of doing patent licensing into three buckets:

1) stick licensing;

2) carrot licensing; and

3) combining patent licenses with something else.

I vaguely recall that he used a less politically correct term for the 3rd approach, but for reasons of decency I will steer clear of it. It begins with a "b". Finally, while things like outright government intervention, invoking familial ties and various other things like that can also create "pull" for patent licensing, I am steering clear of all of those. As they are not specific to patents and some use them to create the "pull" for things ranging from cars to houses to student loans to fine art and to whatnot. So, no weird stuff and let's stick to the legal and businessy spectrum of creating the "pull". In other words, so only working with Marshall's wise classication and expanding it just a little.

STICK LICENSING

This really works. And essentially involves using litigation, or at least a credible threat of litigation, to create the "pull". This is also the hardest and most arduous approach. And carries high risks. Many licensing programs and careers have crashed and burned through this. Intensive. Hard. Adrenalin pumping, but also what has lead to most of the biggest licensing success stories out there. What you essentially do is make a justified and commercially sane demand, a FRAND one if necessary, and then are willing and able to back it up with litigation, if or rather when the other party just stalls. I have so many great memories from stick licensing. This is the way the three out of four largest licensing programs have built their licensing businesses. And many patent trolls / NPEs / PAEs (pick a term you like) have tried this, but almost all have failed. Because it is so hard. And for it to work you need genuine respect and a good innovation story. And you also need to act responsibly, even if you also act firm. Hard to balance that, but table pounding does not work as even stick licensing is very much a people business.

VARIATIONS OF STICK LICENSING

I will add a few variations to Marshall's classic stick licensing. In his defense, they did not really exist when he was most active in the industry. The first one is divesting patents in a limited, but regular, manner to create more "pull". A perfectly legitimate strategy. What you do is you prune and divest some subset of your patents you no longer need or can afford to maintain. And make that a regular process. Then you tell the prospective licensees that, if they stall and refrain to take licenses, they will also have to deal with others having acquired the patents from you. Divesting patents is something commonly and routinely done by virtually all the large patent holders, but it can work to strengthen the "pull" particularly against the companies stalling negotiations and is perfectly legit. And easily explainable to the ones stalling you on licensing discussions. But, importantly, this is just sort of a slight variation of classic stick licensing.

Another variation is that you sell all or substantially all your patents to someone else hoping that they can monetize for you, while you could not, and then you get a revenue share from them. Or whatever. This generally does not work as you no longer have the patents to license and enforce. And the entity having acquired the patents generally does not have good enough knowledge about the patents, access to inventors, any sensible innovation story and so on. All patent litigators will understand what that means. Failed examples of this include entities like Rockstar Consortium and Pendrell/ContentGuard. But I could go on and on. Really struggle to think of any successes with that.

CARROT LICENSING

This involves being the nice guy. And, as is normal in life, it really does not work. Patent litigation is expensive. Think like MUSD 3-6 per lawsuit in the USA, or like MUSD 8-15 for more complex processes before the US ITC. Cheaper, and frankly vastly better value, in places like Germany, but still expensive. So, if you ask very little for your patents, normally people will read that as you not being a serious risk. Your ask may not even cover litigation costs. And they will often have many others chasing them. While I could go on, I stop here. Being overly nice and thinking that super low pricing somehow creates massive "pull" for patents is generally destined to fail. There is one exception though. If you can combine carrot licensing with something else, such as SW, specs or perhaps branding, then yes they together can work to some degree, if the overall value prop makes sense.

And here I would remind that the relative level of patent roaylty demands should always be assessed in relation to the portfolio in question. For some very weak patent portfolios a really low royalty would be the objectively correct royalty (the Motorola H.264 and WLAN portfolios spring to mind). Only problem being that such a portfolio is sometimes not effectively licensable either. Better to keep it for defensive use or sell it to someone larger.

VARIATIONS OF CARROT LICENSING

Again, I would like to add a few variations to Marshall's carrot licensing category. Collective carrot licensing happens through patent pools and is based on the notion that, if you could license lots of patents from lots of entities for an (allegedly) low price, you would be tempted to do so. But it generally does not work. I will later write an article about patent pools and why they are destined to fail, but will spare you for now. All I will say that the success of the MPEG-2 patent pools back in the day was a fluke that nobody has been able to really replicate and that the whole patent pool phenomenon peaked in like 2002 - 2004 or so. The biggest reason being that patent pools are destined to fail on game theory alone, as anyone having spent time in patent pool meetings knows. Further confirmation is available in the form of all the academic articles about patent pools in about 2007 - 2009 time frame. Academic research normally peaks like about 5 years after the underlying phenomen peaked in real life importance. And same happened more recently with FRAND. So, wait for a separate article, but, no, patent pools do not generally work. Then again, never say never, and things my change. Though I doubt it.

There is, however, another variation of carrot licensing that can work. But it is not really an independent one. It involves a situation where a genuine workable licensing program has been built through honest stick licensing. Once the patents in the underlying portfolio start to expire, they may no longer justify a higher royalty rate. But everyone has respect for the portfolio and are used to paying for it. Shifting gradually more over into carrot licensing with lower rates can be a graceful way to end-of-life a formerly active and energetic stick licensing program. Nothing wrong with that. Besides, the F&C types will love having no risk and lower fees for something that used to be a high risk and more pricey. Win-win. Everyone looks good.

COMBINING PATENTS WITH OTHER THINGS

Several licensing programs, including the largest one in the world, have been created by using something else to create the "pull" for patent licenses. Ideally a monopoly or a quasi-monopoly. I will not sell you my monopoly products, unless you take a license. Easy math for many in the short run, if you must, or at least feel you must, get access to said products. I am not assuming any moralistic or legal view on this here. Have my peace with this topic for now, after ... well ... after a lot of stuff. From patent licensing perspective, this obviously works to create the "pull", but importantly the "pull" is not created with patents, but with something entirely else instead. So, you know, I leave it to you whether this is even really patent licensing at all. But it works to create licensing revenue, for sure.

VARIATIONS OF COMBINING PATENTS WITH OTHER THINGS

You can also combine patent licenses with other things like access to source code, other technology implementations, documentation, trade secrets or even product support. The classic example being the pharma industry, where patents are licensed in a big package together with the underlying research, clinical studies, FDA approvals, trademarks and even production knowhow. Another thing that is commonly use on the side to create "pull" for patent licensing is, paradoxically, selling patents to the licensee. Often works, because you are dealing with IP people who both license in and buy patents. So, if they are active on the buy side of the patent market, why not throw in a few to sweeten the license? It is, of course, easy math that overdoing that will lead to you running out of patents. Anyhow, nothing wrong about combining other things with patent licenses. Can work beautifully, but again entails using extensively other things than patents to create the "pull" for patent licenses. Not purely patent licensing. But can work to support stick licensing or even make some forms of carrot licensing feasible.

Finally, there is a third variation of combining patents with other things. This involves using the "pull" of an existing and established patent licensing program to sell other products, perhaps in less demand. This has been and can be done. But is not really an independent approach as it requires an earlier successful era of really stick licensing that has lead to a credible and real licensing program. Which then, for some reason, you want to convert over time to a different approach, where you give up some of the royalties in exchange for the licensee doing something you'd like them to do. Like buying some other products. Or making their products interoperable with yours. Or anything else that is of value to you. Again, works, but risks eroding the earlier established value proposition for the patents in question over time. So, not really an independent approach and perhaps best suited for end-of-lifing a licensing program that was very successful earlier.

CONCLUSION

So, that's it. Made Marshall's 3 approaches into 9 approaches by adding some granularity based on real life insights. But importantly did not break the integrity of his classication. Respect its wisdom too much to break it. Did not venture outside and blabber something about open source or open innovation. You know, I have massive respect for open source and even vividly recall Linus Torvalds himself being (always) logged in on slow modem call-in lines on the Unix servers of the University of Helsinki in the early 1990s. If you did ping him, he would stop whatever he was doing and help you over IRC with whatever Lync, Rtin, Emacs or other problems you had. Obviously that whatever else he was doing was programming Linux at the time. Sorry to digress, but as a former Unix geek his help was invaluable to me. Still thankful.

But the point is that, while real and relevant, open source and open innovation are not really ways to monetize or create the "pull" for patents. To make a massive understatement. You can finesse yourself around certain things there, but I will resist the urge to get into more details. Anyhow, they cannot be classified as ways of creating "pull" for patent licenses.

So, here are the 3 X 3 ways all summed up:

1a) Stick licensing: in its classic form, the basis for everything. Works. Note that this does not mean that it is easy.

1b) Stick licensing with limited scale divestments to increase the "pull": Works to a degree and is explainable to prospective licensees. But requires classic stick licensing too. Also, just business as usual for many patent holders.

1c) Stick licensing with large scale divestments involving the entire or substantially the entire portfolio: Does not work. See, e.g., Rockstar.

2a) Carrot licensing in its classic form: generally fails. People will not generally pay attention to you. Back to the drawing table. You could perhaps make this work by combining the patents with something else (see 3b).

2b) Carrot licensing in the collective patent pool form: fails. On game theory. Wait for a separate article.

2c) Carrot licensing to end-of-life an earlier stick licensing program: works, but is not an independent approach. Requires earlier classic stick licensing.

3a) Combining patents with something else, ideally a monopoly, which creates the "pull": works, but is not really patent licensing.

3b) Combining patents with something else, short of a monopoly, which creates at least some of the "pull": works, but needs help from something else that are not patents. Can be used to support classic stick licensing or even make some forms of carrot licensing feasible.

3c) Combining patents with something else, so that the "pull" of patents is used to sell something else: works, but is not an independent approach in that it requires an earlier stick licensing program. And does not increase the "pull" for the patents, but instead risks eroding it. Because you are using the "pull" for the patents to push some other products.

There you have it. Only one real approach to create the "pull" for patent licensing that is both a) independent of other things; and b) does not require some other approach to have succeeded earlier: is option 1a = classic stick licensing. The only independent form of true licensing of patents. The hardest, but also potentially the most rewarding type. And the one I love. Realize also that this shows how crucial patent litigation is for licensing. It really is a form of marketing and the whole system is premised on it. If you struggle to get comfortable with the key importance of patent litigation, then you will struggle with patent licensing.

Over and out for now. In the meantime, check out 23andme.com.